An efficient disease prediction system using feature optimization and clustering techniques on high dimensional data

Abstract

Human health is very important in this world today due to the rapid newlinechanges of food culture and the reduction of physical activities. so that human newlinehealth is playing major role in all fields. The individual human health is a newlinefundamental requirement today for the developing countries to create a newlinehealthy and wealthy society. As per the World Health Organization (WHO) newlinereport, billions of people are died due to the lack of awareness about the newlinevarious new and old diseases as well. For this purpose, many medical expert newlinesystems and disease prediction systems have been developed by many newlineresearchers to assist the physicians in the direction of decision making and the newlinepublic to get alert about the diseases. Recently, the various technologies newlineavailable like Internet of Things (IoT) to gather the necessary data which is newlinehelpful for making effective decision on patient records that are provided as newlineinput to the disease prediction system. For the purpose of decision making newlineprocess on patient records, the Machine Learning (ML) and Deep Learning newline(DL) were used in the existing disease prediction system. In addition to that, newlinefew meta-heuristic techniques are used to select the required and important newlinefeatures (Symptoms) and also used clustering methods to gather the relevant newlinepatient records. This research work proposes a new disease prediction system newlineto predict the diseases including cancer, heart, diabetic and Arrhythmia by newlineanalysing the patient records by applying the newly developed feature newlineselection and optimization techniques, clustering techniques and deep newlineclassification algorithms. newline